Clinical and biologic studies of geriatric depression are reviewed to evaluate the role of age of onset as a distinguishing parameter. The findings suggest that late-onset depressives have a different presentation, more frequent relapses, and a greater association with medical disorders, dementia, and aging-related biologic changes than early-onset geriatric depressives. Careful clinical evaluation and follow-up of late-onset depressives are suggested. Longitudinal investigations of geriatric depressed patients are needed in order to study the frequency and time of occurrence of various clinical outcomes and identify predictors of outcome in late- and early-onset geriatric depression.